Serving the homeless: evaluating the effectiveness of homeless shelter services.

Fischer, Robert L.

The effects of homeless assistance services at the local level are tremendously difficult to ascertain. In this study, a four-month sample of homeless persons served by a local homeless shelter and case management program were contacted nine to eleven months after receiving services. The findings suggest that the program had some initial success in assisting the homeless clients to locate housing within the first year after leaving the shelter. However, the housing costs paid by these formerly homeless were quite high, with nearly three-quarters of them spending forty percent or more of their income on housing.

**********

Homelessness continues to be a major social issue facing the United States. Depending on the criteria used to operationally define homelessness, the national incidence of the problem has been estimated to range from a low of 300,000 homeless persons to a high of 3.5 million homeless persons (Cordray & Pion, 1990; Rossi, P., Wright, J., Fisher, B., & Willis, G., 1987). In all, an estimated 34 percent of homeless service users are members of homeless families, and 23 percent are minor children (Interagency Council on the Homeless, 1999).

Policy Context

While the multiple causes of homelessness can be attributed in part to the scarcity of low-income housing and the inadequacy of income supports for the poor, clearly there are specific groups of homeless persons who are in need of special services (Burt, 1999; Jencks, 1994; Rossi, 1989; Rossi, 1994). These groups include those with chronic mental illness, alcohol and drug abusers, persons with HIV disease, and families with small children (Cohen, 1989; Cohen & Burt, 1990; Fischer, 1989; Homes for the Homeless, 1998; Lamb & Lamb, 1990).

While the debate over the principal causes of has continued several key findings have been identified. First, there is a persistent group of the poorest members of the population, and among the poorest are children, with some 13 million living in poverty in contemporary America (A. Johnson, 1989) and an estimated 1.5 million homeless youth age 12-17 each year (Ringwalt, Greene, Robertson, & McPheeters, 1998). Fifty percent of African-American children and forty percent of Hispanic children live in poverty, and the single-parent African-American family constitutes the fastest-growing segment of the nation's poor and homeless populations (A. Johnson, 1989). Second, the number of African-Americans who are homeless is disproportionately higher than the percentage of African-Americans in the general population in this country. It has been estimated that, nationwide, nearly 60 percent of all homeless persons are African-American (Homes for the Homeless, 1998), while statistics from the metropolitan Atlanta Area indicate that approximately 80 percent of all local homeless persons are African-American (Atlanta Task Force for the Homeless, 1992). Third, the gap in available housing for the poor versus the number of households in need of low-income housing has widened. In 1993, an estimated 10.6 million units of low-income housing were available for 14.3 million households (Low Income Housing Information Service, 1988). Between 1995 and 1997, the number of affordable units available to low-income households nationwide dropped from 44 units per 100 families to 36 units per 100 families (U.S. Department of Housing and Urban Development, 2000).

Prior Evaluations of Homeless Services

The amount of research devoted to evaluating programs aimed at preventing or remedying the problem of homelessness is exceeded by the numbers of purely descriptive or qualitative studies (Blankertz, Cnaan, & Saunders, 1992; Johnson, & Cnaan, 1995). Some recent work has focused on services for particular categories of the homeless, for example, the homeless mentally ill (Caton, Wyatt, Felix, Grunberg & Dominguez, 1993; Segal & Kotler, 1993), and homeless families (Fischer, 2000; Rog, Holupka, & McCombs-Thornton, 1995; Rog, McCombs-Thornton, Gilbert-Mongelli, Brito, & Holupka, 1995). In addition, the challenges of conducting research with homeless and formerly homeless clients continues to be examined (Orwin, Sonnefeld, Garrison-Mogren, & Smith, 1994). Overall, the existing research on housing outcomes of homeless shelter services consists of primarily small-scale samples of clients, obtained from single communities, and with considerable attrition in the sample at follow-up.

Program Context

Homeless shelter services in northeast Georgia have expanded considerably during the last two decades. In 1974 only four shelters for homeless persons could be found in the metropolitan Atlanta Area, whereas presently approximately one hundred shelters are available (Atlanta Task Force for the Homeless, 1992; Research Atlanta, Inc., 1997). In the local area of Athens, Georgia, during the period of this study, over 3,300 persons were at risk of living in the streets, in shelters, and in overcrowded living circumstances, and thirty-six percent of persons that stayed in area shelters were children under the age of eighteen (Glisson, 1992). The primary local shelter for the homeless in Athens, Georgia, is the Athens Area Homeless Shelter (AAHS). The AAHS placed over 250 persons into permanent housing through its case management services and shelter program during the year in which this study was conducted. However, follow-up information on formerly homeless persons regarding the "durability" of these placements has not been available. An exploratory program evaluation of the AAHS was undertaken in an attempt to determine what happens to the former clients of the homeless shelter after they leave the facility.

Method and Procedures

The present study involved an effort to evaluate homeless shelter services at the local level. The research included a review of program case records and a post-program follow-up with a sample of formerly homeless individuals.

Homeless Shelter Site and Program Services

The Athens Area Homeless Shelter (AAHS) was established in December 1986 and can accommodate up to 32 individual homeless persons in a dormitory-style arrangement, with separate dorms for men and women. Parents and their children can reside in three separate private rooms, each sleeping up to persons.

The intervention used with these homeless individuals consisted of a comprehensive set of services including physical shelter, meals, employment counseling, case management services, supportive counseling, health care referral, clothing supply, and other social services. The primary goal of the AAHS program is to assist clients in obtaining safe, affordable and relatively permanent housing following their departure from the shelter. Secondary goals include assisting shelter clients to obtain employment and to improve their health through proper nutrition and medical care.

Research Design

The base client sample consists of all persons who received AAHS residential services during a four-month period (June to September, 1991), and had a history of residing in the vicinity of Athens, Georgia. Attempts were made to contact all these individuals by telephone or by personal interview approximately 9-11 months following their departure from the Shelter.

A one-page semi-structured interview protocol was developed to assess the following aspects of the lives of former AAHS clients: respondent's current living situation, living costs, length of time at current address, employment and income, perceptions about the safety of their home, and views regarding the AAHS services they had received. The post-test-only design used in this evaluation enables a determination as to what happened to former shelter clients, but not an unambiguous attribution of causation for any positive outcomes, due to the lack of controls inherent in such a research design. Nevertheless, since the AAHS (and most other homeless shelters) had little systematic information on the housing disposition of their clients after they left the shelter, the present inquiry was seen as a valuable first step in documenting the possible outcomes of shelter services.

Survey Results

For the purpose of this study, the unit of analysis is a "head-of-household" and represents either an individual person who sought shelter services solely for him/herself, or the head of a family (e.g., a husband/wife, or a single parent with one or more children). A total of 124 households (individuals or heads of families accompanied by family members) representing 166 men, women and children had received residential services during the sampling time frame. Based on client records maintained by the AAHS, at entry into the program, 75 percent of the sample group's earnings were below the federal poverty line for the relevant size of household groups. Sixty-six percent of the sample group members were African-American, 30% were white, 4% were Hispanic, and less than one percent was Asian. Although African-Americans make up the majority of homeless persons served at the AAHS (as well as of our sample group), African Americans comprise only approximately one-quarter of the general population in the Athens metropolitan area. The sample groups' stay in the AAHS averaged nearly three weeks, but ranged from one night to six months.

Of the 124 households, intake information indicated that 100 households (81%) had a history of residing in the Athens vicinity. The researchers with the assistance of AAHS staff sought out these individuals and families for the purpose of conducting a follow-up interview. The follow-up efforts resulted in contact with 71 of the 100 Athens-resident households (71% response rate) for follow-up interviews. The remaining 24 households were not contacted due to a lack of information in their client file and were unable to be traced. Thus, the housing circumstances of these 24 nonrespondent households are unknown. However, a follow-up contact rate of 71% is a substantially higher than would be expected, considering the nature of homelessness (A. K. Johnson, 1989).

At follow-up, which ranged from 9-11 months (average of 38 weeks) following the client's departure from the AAHS, 41 of the 71 former clients (58%) held contractual agreements (i.e., leases) in their own name, indicating that they occupied relatively permanent housing. The former clients had, on average, resided in their current home for eighteen weeks, with a range from one week to 50 weeks. Some of the housing characteristics and living circumstances of the former AAHS clients are presented in Table 1.

Nearly a third (31%) of the clients had maintained their follow-up home for six or more months, and 35% had lived in their home the entire period of time since leaving the AAHS.

Data on monthly housing costs were obtained from 51 of the 71 former clients (72%); the remaining 20 persons did not provide this information or had no direct housing costs, such as instances in which the individual was living with a relative. For these 51 respondents, their monthly rent averaged $186 and utilities averaged $47, for a mean total monthly housing cost of $233 (range = $50 to $645). Monthly income for the 51 clients reporting housing costs averaged $503, and thus these individuals expended approximately 46% of their monthly income on housing. Of the 51 households reporting, 39 households were residing in private housing and 12 households were in public housing. Overall, 20 households' (28%) percentage of monthly income spent on housing costs was over 50 percent. Surprisingly, 6 of the 12 households living in public housing (50 percent) reported spending 50 percent or more of their income on monthly housing costs. Although the clients' rents were set at 30 percent of their monthly income, utility costs pushed half of these persons' monthly housing costs above the 50 percent threshold. Additionally, 14 of the 39 private housing households (36%) reported monthly housing costs of over 50 percent of one's monthly income. Sixty of the 71 respondents provided information about their monthly income, and according to these data, 55 of the 60 households (92%) earned less than the federal poverty standard ($9,100 for a single individual in 1992; $13,700 for a family of three). For the African-American households reporting housing costs, monthly housing costs consumed 50% of their monthly income, while for the white households these costs amounted to 39% of their income. White and African-American households had entered into contractual housing agreements (i.e., leases) in approximately the same proportion (60%). African-American households moved fewer times (M = .88 times) than white households (M = 1.4 times) and, on average, the 51 African-American households had lived in their present living arrangements for 18 weeks while white households had done so for 16 weeks.

One key dimension of a desirable housing situation for homeless individuals and families is the level of safety. Overall, the clients' average perceived safety rating of their present home was 3.0 (O.K.), but a difference was present in that those living in public housing (n = 12) provided an average safety rating of 2.3 while that for the 32 respondents residing in private housing gave a mean rating of 3.2. Clearly former clients living in public housing projects felt less safe in their home and environment than those in private circumstances.

Discussion

The findings presented here can be cautiously seen as positive. A majority of the respondents (58%) were residing in stable housing situations at the time of the follow-up interview, and nearly half (45%) had resided in their current dwelling for about four months since their departure from the homeless shelter. On average, the formerly homeless persons rate the safety of their homes as acceptable.

In relative terms, the housing costs of these formerly homeless individuals and families are high: 36 of 51 respondents (71%) spend 40% or more of their income towards housing costs, well above the U.S. Department of Housing and Urban Development's standard of 30% as an appropriate proportion of one's income which should apply towards housing. Only 6 of the 51 respondents (12%) paid less for housing than this federal standard. Thus, it would seem that most of the AAHS clients remained at-risk in terms of their ability to maintain a stable home situation, given the large proportion of their income going towards housing costs. After a short stay in the homeless shelter, the majority of these individuals are able to locate relatively safe, affordable and stable homes. It should be noted, however, that the majority of the AAHS former clients continued to live in poverty and perilously on the verge of a return to homelessness.

Table 1
Housing Environment and Living Circumstances of Shelter Clients
Contacted at Follow-Up (n = 71)

Housing Type at Follow-up
 Apartment 36.6%
 Mobile home 15.5%
 Single-family home 12.7%
 Duplex 8.5%
 Rented room 8.5%

Living Circumstances at Follow-up
 Living alone 18.3%
 Living with relatives 16.9%
 Living with own child(ren) 14.1%
 Living with friend(s) 12.7%
 Living with boy friend 9.9%
 Living with spouse 5.6%
 Living with roommate 4.2%
 Other circumstances * 18.3%

* includes living on the streets, in a shelter, etc.

References

Atlanta Task force for the Homeless. (1992). Statistical Report for 1991. Atlanta, GA: Author.

Blankertz, L. E., Cnaan, R. A., & Saunders, M. (1992). Assessing the impact of serving long-term mentally disabled homeless persons. Journal of Sociology and Social Welfare, 19, 199-220.

Burt, M. R. (1999). Demographics and geography: Estimating needs. In Fosburg, L. B., & Dennis, D. L. (eds.), Practical Lessons: The 1998 National Symposium on Homelessness Research. Washington, DC: U.S. Department of Housing and Urban Development and U.S. Department of Health and Human Services. 1:1-1:24. August.

Caton, L. M., Wyatt, R. J., Felix, A., Grunberg, J., & Dominguez, B. (1993). Follow-up of chronically homeless mentally ill men. American Journal of Psychiatry, 150, 1639-1642.

Cohen, M. B. (1989). Social work practice with homeless mentally ill people: Engaging the client. Social Work, 34, 505-509.

Cohen, M. B., & Burt, M. R. (1990). The homeless: Chemical dependency and mental health problems. Social Work, 26, 8-17.

Cordray, D. S., & Pion, G. M. (1990). What's behind the numbers? Definitional issues in counting the homeless. Housing Policy Debate, 2(3). 587-616.

Fischer, P. J. (1989). Estimating the prevalence of alcohol, drug, and mental health problems in the contemporary homeless population: A review of the literature. Contemporary Drug Problems, 16, 333-390.

Fischer, R. L. (2000). Toward self-sufficiency: Evaluating a transitional housing program for homeless families. Policy Studies Journal, 28(2). 402-420.

Glisson, G. M. (1992). Athens Area Homeless Service Providers Report, 1991. Athens, GA: Athens Area Homeless Shelter.

Homes for the Homeless. (1998). Ten Cities: A Snapshot of Family Homelessness Across America. New York: Author.

Interagency Council on the Homeless. (1999). Homelessness: Programs and the People They Serve. Washington, DC: U.S. Department of Housing and Urban Development. December.

Jencks, C. (1994). The Homeless. Cambridge, MA: Harvard University Press.

Johnson, A. (1989). Poverty and homelessness. Public Welfare, 47(1),29-42.

Johnson, A. K. (1989). Measurement and methodology: Problems and issues in research on homelessness. Social Work Research and Abstracts, 25(4), 12-20.

Lamb, H. R. & Lamb, D. M. (1990). Factors contributing to homelessness among severely mentally ill. Hospital and Community Psychiatry, 41,301-305.

Low Income Housing Information Service. (1988). Housing in America. Washington, DC: Author.

Orwin, R. G., Sonnefeld, L. J., Garrison-Mogren, R., & Smith, N. g. (1994). Pitfalls in evaluating the effectiveness of case management programs for homeless persons: Lessons from the NIAAA Community Demonstration Program. Evaluation Review, 18(2). April 153-207.

Research Atlanta, Inc. (1997). Homelessness in Metro Atlanta. Atlanta: Georgia State University, School of Policy Studies.

Ringwalt, C. L., Greene, J. M., Robertson, M., & McPheeters, M. (1998). The prevalence of homelessness among adolescents in the United States. American Journal of Public Health, 88(9). 1325-1329. September.

Rog, D. J., Holupka, C. S., & McCombs-Thornton, K. L. (1995). Implementation of the Homeless Families Program: 1. Service models and preliminary outcomes. American Journal of Orthopsychiatry, 65(4). 502-513. October.

Rog, D. J., McCombs-Thornton, K. L., Gilbert-Mongelli, A., Brito, C., & Holupka, C. S. (1995). Implementation of the Homeless Families Program: 2. Characteristics, strengths, and needs of participant families. American Journal of Orthopsychiatry, 65(4). 514-528. October.

Rossi, P. H. (1994). Troubling families: Family homelessness in America. American Behavioral Scientist, 37(3). 342-395. January.

Rossi, P. H. (1989). Down and Out in America. Chicago, IL: University of Chicago Press.

Rossi, P., Wright, J., Fisher, B., & Willis, G. (1987). The urban homeless: Estimating composition and size. Science, 234, 1336-1371.

Segal, S. P. & Kotler, P. L. (1993). Sheltered care residence: Ten-year personal outcomes. American Journal of Orthopsychiatry, 63, 80-91.

U.S. Department of Housing and Urban development. (2000). Rental Housing Assistance: The Worsening Crisis. Washington, DC: Office of Policy Development and Research. March.

GEORGE M. GLISSON
BRUCE A. THYER
University of Georgia
School of Social Work

ROBERT L. FISCHER
Families First